194 SIR DAVID BREWSTER ON THE STRUCTURE AND 



Even when the glass has been exposed only to the air of the atmosphere, 

 these coloured films are produced sometimes in twenty or thirty years on the 

 surface of glass, whether formed by fusion or by artificial grinding and polishing. 

 But as other surfaces of glass, similarly exposed, and of a much greater age, show 

 no marks of decomposition, it seems evident that the rate of decomposition must 

 depend upon the composition of the glass, and probably on the temperature at 

 which its elements have been combined. 



A very remarka])le case of rapid decomposition presented itself to me in a 

 fine plate-glass prism, made for Mr Talbot by Fraunhofer at Munich, and 

 presented to me by that distinguished philosopher. On the 23d March 1833 I 

 made accurate drawings of its three faces, on each of which there was a circular 

 spot, in which the decomposition showed itself in three rings or orders of the 

 colours of thin plates. In one of these spots, half of which is cut off" by the edge 

 of the prism, the colours in 1833 were three orders, the innermost being a 

 yellowish-green of the third order. This spot is about a quarter of an inch in 

 diameter. Another spot about the same size is perfectly circular, the outermost 

 ring being the white of the^r^^ order, rising in the centre to imrplish-blue of the 

 second order. The third spot, about the 16th of an inch in diameter, is perfectly 

 circular, the innermost tint being a red of the second order. These colours, which 

 are very difficult to be seen, are those observed at the polarising angle of the 

 glass, when the reflected light is nearly extinguished by an analysing prism. 



Upon examining these decompositions after an interval of thirty years, I can- 

 not observe any change in the rings and colours. They seem to be fainter than 

 before, and, what is very remarkable, two long irregular streaks of decomposition, 

 one an inch long, and the other nearly two inches, have entirely disappeared. 

 In the shortest of these lines of decomposition, the tints of some parts rose from 

 blue to yellow in the longest, and in a very small circular spot the tint was 

 only blue. 



The pause in the process of decomposition in the three circular spots, and the 

 disappearance of the long streaks, is very remarkable, — and the more so, as they 

 must all have been formed during a period certainly not greater, but probably 

 very much less, than twenty years. Is it not possible that the particles which 

 have separated from the glass by the action of the decomposing cause, and 

 ceased to be in optical contact with it, may have returned into optical contact 

 when the decomposing cause no longer existed, as might have been the case 

 with the prism if kept, as it has been, in dry air since 1833 ? That this is not 

 an unreasonable supposition, may be inferred from an experiment published in 

 1816, in which one of the surfaces of a fissure made artificially in a thick plate 

 of glass was so much separated from optical contact with the other surface, as 

 to reflect totally light incident upon it at a certain angle. " After standing 

 an hour the fissure began to disappear, and in the course of a day it was as 



